Yet in

gle instance in which the male is larger than the female." the very chapter containing this assertion Darwin represents three species of fishes selected by Günther in which males are figured of larger size than females; they are named Callionymus lyra (fig. 29), Xiphophorus helleri (fig. 30), and Plecostomus barbatus (fig. 31). These species, be it remarked, have the sexes trenchantly differentiated, the males in two cases being marked by a superior brilliancy of coloration and exuberance of fins and in the other (Plecostomus) by a bristly armature of the head. The instances of increased size of the male are, in fact, numerous; in almost all cases when the males are decidedly differentiated from the females by brilliancy of color or other secondary characters the rule is that they are larger than the females. Like other rules, this may be subject to exceptions, but the rule is based on extensive observations. A couple of figures made

simply to illustrate sexual differences for Mr. C. Tate Regan's Monograph of the Loricarioid Catfishes of America and reproduced here exemplify the rule. Of a kindred species, Pseudancistrus barbatus," the sexes are illustrated in Darwin's work.

During the consideration of the social economy of these fishes the question must often recur, How did the parental instinct manifested originate? It was easy enough in olden time to give an answer which would be regarded as all sufficient in those days; it was a specific instinct implanted by an omnipotent creator in every case. In these days of evolutionary belief, however, such an answer is equivalent to no answer. The instinct must be regarded as a development of an aptitude inherent in the fish itself.

a Darwin, The Descent of Man, N. Y., 1881, p. 335.

↳ Plecostomus barbatus Günther, Darwin; Ancistrus barbatus, Regan, SM 1905 30

Probably few naturalists or psychologists will be prepared to concede the possession of a sentiment of antiselfish altruism by any of these lowly forms. The attribute of parental care must therefore be regarded as an outcome of selfishness, or, if you will, self-love," a result of the sense of proprietorship. The eggs are the fish's own, and therefore they and the resulting larvæ are to be cared for as such. Perhaps it may be urged that the attention of the parental fishes is of the same nature as that of the hen to her young. We are not prepared to deny it. It may even be conceded, and yet the claim that the sentiment is the offspring of self-love can still be maintained. In fact, there is a regular gradation of self-love into the ennobled sentiment which impels the human mother to sacrifice her life cheerfully for her child and the degraded passion which emboldens the miser to

FIGS. 4, 5.- Ancistrus chagresi. After Regan.

suffer death rather than lose his gold. It is the basis of the courage of the farseeing martyr for his religion, for he is willing to sacrifice the present for an illimitable future.

Wonder may be entertained that one and the same method of care should have originated independently many times, but this will diminish on reflection. When the sense of proprietorship in the eggs has been established protection by hiding them or clearing away of foreign substances that would interfere with them would not unnaturally follow. The mouth is used by many fishes for carrying, and the instinct to take up the eggs into the mouth for protection would be a natural consequence which might be, and repeatedly has

a See Whately's Morals and Christian Evidences, XVI, see. 3 (Am, ed., 1857, p. 129); Ward's Dynamic Sociology, I, 679.

been, developed into a habit. These and other provisions for the care of the eggs do not make excessive demands on our receptive capacity or imagination. It is only when we consider the case of the sticklebacks that the combination of aptitudes for nest making impresses us. In them complexity is carried to an extreme. There is a sympathetic development of the kidneys and the testes; there is the synchronous response of both to external stimuli; there is the reaction of both on the brain and of the brain to external conditions; there is the elaboration of the wonderful thread which is used to bind the nest materials; there is the instinct to use the thread; there are, finally, the regular aptitudes and impulses which are shared with the majority of fishes. Such an accumulation and convergence of structural, physiological, and psychological characters almost force upon us a rejection, as explanation, of natural selection or sexual selection. The development manifested in the Gasterosteids is, indeed, one of the greatest wonders of the evolution of animal life. Nevertheless, it may not seem so extraordinary if we extend our researches beyond the class to which they belong.

Naturally the elaboration for the perpetuation of the species in the Gasterosteids is unique in the class of fishes, but there is to some extent an analogous provision, although not so complicated, for the animal economy in certain other classes, as in the case of the webs of most spiders and the cocoons of some insects, as notably the silkworms. Mollusks and Annelids furnish other illustrations. Of course in those cases the analogous secretions are not produced by homologous organs, but by very different ones, very different from each other as well as from those of the sticklebacks.

As this article is intended to summarize existing knowledge respecting the breeding habits of some of the fishes in question, the original words in which the facts are recorded are given in most cases in fact, wherever it could be done without interference with the mode of treatment or sequence adopted for the narration. The text, however, is by no means confined to information respecting the nests and breeding habits. Most of the species are little known or even quite unknown to many persons and yet are very interesting for other reasons than their parental instincts. Consequently quite detailed accounts of their habits in general as well as their places in nature are added. The illustrations are mostly borrowed and are derived from various sources, as will be indicated in connection with the respective figures. Among the original illustrations are those of a fish longest known-the Glanis of the ancient Greeks, never before figured.

THE DIPNOANS.

was

In and confined to the Southern Hemisphere are three very peculiar types of aquatic animal life. One from South America was first described in 1836 as a reptile related to the siren of the north and named Lepidosiren; another, from Africa, not long after (1841), made known as a true fish, and on account of its simple and supposedly primitive fashion of limbs styled Protopterus. For a long time it was a matter of dispute between naturalists whether the two were reptiles or fishes, and no relationship was recognized between them and any other forms, recent or fossil. At last, in Australia, in 1870, was discovered the third type, and then it became evident that not only had they all relatives in the past, but all their relatives were of the past, and the very distant past. The most recent of those extinct forms, so far as known, lived not only in old Europe and America, but also in Africa, Asia, and Australia. They enjoyed a world-wide distribution during the Jurassic period and thus early in geological history disappeared from the surface of the earth. Some, of course, must have survived to transmit their blood and likeness to

FIG. 6.-Neoceratodus forsteri. After Günther.

their living relatives, but their fossil remains have not yet been found. There is an extraordinary gap between the oldest of the living types (of very late Tertiary age) and the hosts that once ranged over the globe. The three types are the Lepidosiren of South America, the Protopterus of Africa, and the Neoceratodus of Australia. They and their distant relative, Polypterus, are immeasurably the nearest of kin to the stock from which alike fishes and reptiles originated. Two are remarkable for the provision which they make for their eggs and young. It is most fitting, therefore, that they should be the first to illustrate parental care among fishes.

The three genera are now segregated into a group named Dipnoi or Dipneusti, and often called lung fishes, which has been variously estimated to be of subclass, ordinal, or even class value. Here the subclass valuation may be accepted as expressing best the taxonomic importance of the distinctive characters of the group. Among living fishes the only ones that are at all related to the Dipnoans are the Polypterids or Bichirs of Africa. These are the representatives of a

great group or subclass named Crossopterygians which were characteristic or even predominant fishes of ancient times from the Devonian to the Jurassic period. One of the most striking external features of the Crossopterygians was the prolongation of the axes of the paired fins and the consequent lateral or fringelike arrangement of the rays along the axes. It is to this characteristic that the name crossopterygian (or fringe-finned) alludes.

During the same remote period in which the ancient Crossopterygians flourished lived also fishes so much like those forms that they were confounded with them in the same group by so good a naturalist as Huxley and by many others. They likewise had lobate paired fins, but later it was found that instead of a distinct suspensorium for the lower jaw there was no distinction between the suspensorial elements and the cranium, and that all formed one piece with which the lower jaw directly articulated. There were no distinct upper jawbones. The heart of the living representatives has the same kind of antechamber as that of the Crossopterygians although otherwise different.

It was a long time before the facts thus epitomized became known and appreciated. The histories of the ancient forms and the recent ones were long told as those of beings entirely unrelated. That of the living ones is as interesting as it is curious.

Two very distinct families of Dipnoans are represented in the modern world, one (Ceratodontids) confined to northern Australian rivers and the other (Lepidosirenids) common to South America and Africa.

In 1836 a great naturalist-collector, Johann Natterer, discovered in the Amazon basin an animal of which he sent two specimens to the Imperial Austrian Museum, and by the custodian of the museum (Fitzinger) it was described as a reptile related to the North American Siren and called Lepidosiren paradoxus. Several years later (1841) Thomas Weir sent two specimens of another animal, taken from the Gambia River, West Africa, to London. This species was described by Owen and considered to be a true fish related to the eels and was at first called Protopterus annectens, but later it was referred by its describer to the same genus as the South American animal. Owen thought the two belonged to a peculiar "family" and made "the nearest approach in the class of fishes to the Perennibranchiate reptiles." In fact, none of the old authors had any real appreciation. of the relations of either animal.

Quite a warfare was carried on for some time about the question whether the animals were reptiles (amphibians) or fishes, but long ago it was decided in favor of the piscine relationship of the Lepidosirenids.